When he is talking about linearity on the order of ppm (parts per million) he has a point - not all that easily achieved even in the world of precision op-amps, but not achievable in any practical sense with tube circuitry at all. (Open loop gains of >1e6 are required to get anywhere close.) And IMHO it doesn't matter - there are other criteria that may be more important, at least to my ears, and the ear is not exactly a linear device, could it just be that the linearity limitations of typical tube circuitry better complement our ears than something that on the theoretical level is far better?
Anyway this is an argument that no one is going to win, and I hope I do not have to spell that out.
All arguments will have their proponents and well reasoned ones may demonstrate a degree of validity despite other equally well reasoned arguments from the other side, whichever it may be.
So my biases are clear I am a tube guy at home and a solid state guy at work.. My digital stuff is good, and not a tube in sight. (Feeding tube amplification) Analog path all tube both tape and LP..
That s not a belief , it s an established fact, backed by any possible
measurement of any parameter that count when defining linearity.
It s not only a matter of THD.
Take any parameter that has an influence on the amplification
accuracy, and inevitably , the (correctly designed) SS amp
will be largely superior to any tube amp.
Wether we talk of phase response , damping factor , output noise ,
bandwith regularity , intermodulation, slew rate , or any other
parameter , the tubes are severly lacking.
That s too much factors at the same time that goes in the same direction.
Of course , some people will undermine and even negate such results
by simply removing not parts , but the whole equation , and then
asking to rely on ears+brain , wich will inevitably give biaised results
as the comparison point will be no more accuracy but "good sounding",
while knowing blatlantly that human ears love the addition of some
harmonics when the music recorded sound is too harsh.
But then, the question is no more about linearity or accuracy , but about
what is the most pleasant distorsion and reshaped phase and frequency
responses....
and like me Kevin......you can't explain that can you ?
Music makes you forget the crap sometimes - doesn't it ?
That's why we all press ' PLAY ' which can sometimes make us relax, grin, cry or want to dance.
My dancing is pretty rubbish but all the others apply occasionally
It's nice to see clever, knowledgeable guys like you say.......' erm.. I like it - but don't ask me to tell you why '
Bravo
Music makes you forget the crap sometimes - doesn't it ?
That's why we all press ' PLAY ' which can sometimes make us relax, grin, cry or want to dance.
My dancing is pretty rubbish but all the others apply occasionally
It's nice to see clever, knowledgeable guys like you say.......' erm.. I like it - but don't ask me to tell you why '
Bravo
You are missing the forest for the trees. And what sort of source do you think you use for calibration?
Individual tubes may be far more linear than a solid state device, but have you measured the performance of an op-amp with an OLG of 3.5e6 with CLG of unity and compared it to a cascade of tube stages trying to do the same thing? It isn't possible. (And I am NOT even saying the solid state device sounds better.)Now, in terms of what tubes bring to audio I am completely on your side. I'm not sure that linearity levels beyond that capable in simple tube stages brings anything to audio reproduction.
Funny that you comment on my familiarity with tubes, obviously you don't know me that well yet..
Dig around a bit more..
and like me Kevin......you can't explain that can you ?
Music makes you forget the crap sometimes - doesn't it ?
That's why we all press ' PLAY ' which can sometimes make us relax, grin, cry or want to dance.
My dancing is pretty rubbish but all the others apply occasionally
It's nice to see clever, knowledgeable guys like you say.......' erm.. I like it - but don't ask me to tell you why '
Bravo
I have some theories, none of which I have been able to prove to my level of satisfaction. I also will profess to prefer the clearly technically inferior LP to anything less than 2496 digital and usually even then.
I just know what my ears appreciate, and have a passing acquaintance with what live acoustical music sounds like, and like to hear something analogous at home. Oh, no now I've really done it ..
I'm talking on a purely theoretical level.
I'm a tube/valve guy through and through..It is happy that the old records were made using a minimalist
number of stages , along with a few tracks magnetic tube
preamplified recorder.
If it was done as currently, putting dozen of stage in the audio chain,
but tube ones instead of SS ones, the final result would be awfull.
It s because today s recordings are made using solid state devices
that produce very little noise and distorsions that you can use
ultimately, as the last amplifying element , a tube amp without
adding too much noticeable noises and distorsions...
Since I have heard both Lumley valve electronics, and stacked Quads I know exactly what I would be taking home, and there is no "N" in it anywhere..
Had an SME 20 for a while, superlative table.. (Loaner, too costly for me to own and maintain.) Currently I have (own) a TD-124/II on slate plinth with Schick arm and a variety of low compliance Denon LOMC carts. It's rather good actually. Sorry for OT.. I have some theories, none of which I have been able to prove to my level of satisfaction. I also will profess to prefer the clearly technically inferior LP to anything less than 2496 digital and usually even then.
I just know what my ears appreciate, and have a passing acquaintance with what live acoustical music sounds like, and like to hear something analogous at home. Oh, no now I've really done it ..
Oh yes you have.............time to put on the crash helmet again
So Evenharmonics, can you share with us what amps your graphs relate to and what topologies they use.
Talking about topologies: since about a year I have been working with a number of class D amps, and I must say I haven't been listening to my other amps since. I let my ears do the voting.
Now, using FETS in the output stage of class D amplifiers makes sense because in SS these devises can cope with the high switching speeds. Is anybody aware of serious approaches to build class D with tube end stages?
It may have a number of advantages. The output transformer would effectively choke the switching frequency. Furthermore, deadband might be reduced with benefits for distortion. First of all because of the high switching speed you can achieve with tubes, second because I suspect tubes will be much more resilient to through shoot than SS devices.
Talking about topologies: since about a year I have been working with a number of class D amps, and I must say I haven't been listening to my other amps since. I let my ears do the voting.
Now, using FETS in the output stage of class D amplifiers makes sense because in SS these devises can cope with the high switching speeds. Is anybody aware of serious approaches to build class D with tube end stages?
It may have a number of advantages. The output transformer would effectively choke the switching frequency. Furthermore, deadband might be reduced with benefits for distortion. First of all because of the high switching speed you can achieve with tubes, second because I suspect tubes will be much more resilient to through shoot than SS devices.
Sadly I was not the owner of either system but those moments you do not forget.
My old man has a single pair of ESL 57's with Quad amps and it still sounds lovely with his old LP12 and yes - I agree - vinyl rocks
Good job you got a sense of humor K - when two tech brains collide - we've got to run for cover
My old man has a single pair of ESL 57's with Quad amps and it still sounds lovely with his old LP12 and yes - I agree - vinyl rocks
Good job you got a sense of humor K - when two tech brains collide - we've got to run for cover
It is happy that the old records were made using a minimalist
number of stages , along with a few tracks magnetic tube
preamplified recorder.
If it was done as currently, putting dozen of stage in the audio chain,
but tube ones instead of SS ones, the final result would be awfull.
It s because today s recordings are made using solid state devices
that produce very little noise and distorsions that you can use
ultimately, as the last amplifying element , a tube amp without
adding too much noticeable noises and distorsions...
Hmm, but I listen to a mix of modern and ancient recordings and my system handles them pretty faithfully regardless of their origins. Wouldn't have it any other way, and I guess that is the point. In fairness I have heard at least one fully solid state system that did things more or less in the same way my system does, and the two systems could not have been more different in design choices and execution, but the goals and care taken were similar as were the results.
The way I see it is that there are many roads converging on potentially the same result, what really matters is the care in design and execution and a clear understanding of the goal. As I get older I realize there is no inherently right or wrong way, just a set of preferences. I prefer tubes, someone else may prefer solid state, comparable goals and skills with either technology may result in convergence.. I've experienced it.
As for preference , that s another debate wich rely on purely
emotionnal choices.
I would agree that ear+brain discrimination capacities are insufficent
to make the difference when comparing correctly implemented designs ,
wether it s tubes or transistors , with some exceptions , as experience
taught me that SS amp really sound more dynamic than tubes ones ,
but when listening at usual volume , it wont be apparent.
Of course, speaker s quality would far more caracterise an audio
chain than any amplifier , provided it s not a total failure...
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.